基于Elman神经网络和遗传算法的油田指标预测#

Forecast of Oilfield Indexes Based on Elman Neural Network and Genetic Algorithm

基于指标预测在油田实际开发中的应用,提出将神经网络和改进的遗传算法结合起来构建预测模型。神经网络采用具有动态反馈的Elman网络,充分发挥其动态预测的优势,同时借助遗传算法弥补其训练速度慢和容易陷入局部极小的缺点。对遗传算法的选择算子加以改进,不仅可以保存优良个体而且可以提高搜索效率。将神经网络和遗传算法进行有机结合,实现优势互补,以大庆葡北油田三断块的后续水驱含水率实测数据为例对模型进行论证,结果表明,该模型能达到很好的指标预测效果,本文提出的方法是有效可行的。

Based on the application of the index prediction in oil field actual development,this paper puts forward a method combining neural network and genetic algorithm build a forecasting model.Neural network using a dynamic feedback Elman network,gives full play to the advantages of its dynamic prediction,and at the same time,by the help of a genetic algorithm compensates for its slow training speed and easy to fall into the local minimum points.By improving genetic algorithm＇s selection operator it not only can save fine individual but also can improve the search efficiency.This paper organically combines neural network with genetic algorithm and realizes the complementary advantages.Taking the oilfield real data for example to test and verify model,the results show that this model can achieve good prediction effect,the method is efficient and feasible.